learning_rate: 0.0008
train_batch_size: 2
eval_batch_size: 2
seed: 42
distributed_type: multi-GPU
num_devices: 8
total_train_batch_size: 16
total_eval_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 12
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
0.7897	0.2001	720	0.6915
0.6597	0.4001	1440	0.6516
0.5927	0.6002	2160	0.6070
0.5746	0.8002	2880	0.5887
0.55	1.0003	3600	0.5604
0.5211	1.2003	4320	0.5440
0.5145	1.4004	5040	0.5423
0.5079	1.6004	5760	0.5268
0.4965	1.8005	6480	0.5160
0.4933	2.0006	7200	0.5142
0.4734	2.2006	7920	0.5031
0.4688	2.4007	8640	0.4952
0.4544	2.6007	9360	0.4921
0.4582	2.8008	10080	0.4807
0.455	3.0008	10800	0.4711
0.4311	3.2009	11520	0.4731
0.4325	3.4009	12240	0.4623
0.4241	3.6010	12960	0.4550
0.4239	3.8011	13680	0.4613
0.4207	4.0011	14400	0.4526
0.3945	4.2012	15120	0.4405
0.4019	4.4012	15840	0.4501
0.3926	4.6013	16560	0.4438
0.3996	4.8013	17280	0.4254
0.3869	5.0014	18000	0.4323
0.3691	5.2014	18720	0.4234
0.3672	5.4015	19440	0.4219
0.3685	5.6016	20160	0.4162
0.3683	5.8016	20880	0.4102
0.3608	6.0017	21600	0.4009
0.3437	6.2017	22320	0.4015
0.3464	6.4018	23040	0.3947
0.3393	6.6018	23760	0.3934
0.339	6.8019	24480	0.3941
0.3324	7.0019	25200	0.3790
0.3168	7.2020	25920	0.3782
0.3135	7.4021	26640	0.3681
0.3138	7.6021	27360	0.3676
0.3086	7.8022	28080	0.3577
0.3154	8.0022	28800	0.3620
0.2904	8.2023	29520	0.3604
0.2869	8.4023	30240	0.3555
0.289	8.6024	30960	0.3453
0.2825	8.8024	31680	0.3440
0.286	9.0025	32400	0.3409
0.2613	9.2026	33120	0.3386
0.2576	9.4026	33840	0.3347
0.2632	9.6027	34560	0.3317
0.2595	9.8027	35280	0.3233
0.2524	10.0028	36000	0.3269
0.233	10.2028	36720	0.3256
0.2324	10.4029	37440	0.3225
0.2305	10.6029	38160	0.3183
0.232	10.8030	38880	0.3152
0.2331	11.0031	39600	0.3097
0.2136	11.2031	40320	0.3084
0.2092	11.4032	41040	0.3078
0.2048	11.6032	41760	0.3056
0.2058	11.8033	42480	0.3064

yalhessi/lemexp-task1-v2-lemma_object_small-Llama-3.2-1B-ddp-8lr-v2

lemexp-task1-v2-lemma_object_small-Llama-3.2-1B-ddp-8lr-v2